Signed-off-by: Ricardo Biehl Pasquali <pasqualirb@xxxxxxxxx> --- man7/socket.7 | 27 +++++++++++++++++++++++---- 1 file changed, 23 insertions(+), 4 deletions(-) Note: I've already sent this patch two times 2018-10-21 and 2018-12-20. Jacob E. Keller gave a feedback in the first send. I've commented some issues of error queue polling in network mailing list: https://www.spinics.net/lists/netdev/msg529722.html Ricardo Biehl Pasquali wrote (2018-10-18): > The commit 7d4c04fc170087119727 ("net: add option to enable > error queue packets waking select") (2013-03-28) introduced > SO_SELECT_ERR_QUEUE, which masks POLLPRI with POLLERR event > return in some socket poll callbacks. > > POLLERR event issued with sock_queue_err_skb() did not wake > up a poll when POLLERR is the only requested event because > sk_data_ready() (sock_def_readable()) was used and it > doesn't mask POLLERR in poll wake up: > > wake_up_interruptible_sync_poll(&wq->wait, > EPOLLIN | EPOLLPRI | > EPOLLRDNORM | EPOLLRDBAND); > > If POLLIN or POLLPRI are requested, for example, poll does > wake up. > > POLLERR wakeup by requesting POLLPRI is possible without > set SO_SELECT_ERR_QUEUE. All the option does is masking > POLLPRI as a returned event before poll returns. poll > would return anyway because of POLLERR. > > Also, the sentence "[...] enable software to wait on error > queue packets without waking up for regular data on the > socket." from the above commit is not true. > > A POLLIN event issued via sock_def_readable() wakes up > threads waiting for POLLPRI, and vice versa. However, > poll() does not return, sleeping again, as the requested > events do not match events. > > The commit 6e5d58fdc9bedd0255a8 ("skbuff: Fix not waking > applications when errors are enqueued") (2018-03-14) make > POLLERR alone wake up poll. It replaces sk_data_ready() > (sock_def_readable()) with sk_error_report() > (sock_def_error_report()). This makes "POLLERR wake up by > requesting POLLPRI" obsolete. > > Rationale: > > POLLIN-only and POLLERR-only wake up are useful when there > is a receiving thread, a sending thread, and a thread that > get transmit timestamps. The thread polling on POLLERR will > not wake up when regular data arrives (POLLIN). The thread > polling on POLLIN will not wake up when tx timestamps are > ready (POLLERR). > > One solution is adding an option that disable POLLERR as > requested event. This is in the Virtual File System > subsystem, not in the network, though. > > This solves the problem of waking up other threads that > not interested in error queue. Thus allowing a separate > thread take care of error queue (useful for receiving > transmit timestamps). > > However, this may not be a good solution as it depends on > polling behavior. Thoughts? > > By the way, as the kernel development shouldn't break user > space, SO_SELECT_ERR_QUEUE can become a compatibility > option. I have tried to explain the effect of the option and the context in which it was introduced (more or less the reason why it was introduced). The POLLIN-POLLPRI mutual wake up issue was not commented. Is it clear? Should a NOTE be written about the other issues? Cheers! pasquali diff --git a/man7/socket.7 b/man7/socket.7 index 478fa364c..1ba308198 100644 --- a/man7/socket.7 +++ b/man7/socket.7 @@ -49,10 +49,6 @@ .\" commit a8fc92778080c845eaadc369a0ecf5699a03bef0 .\" Author: Pavel Emelyanov <xemul@xxxxxxxxxxxxx> .\" -.\" SO_SELECT_ERR_QUEUE (3.10) -.\" commit 7d4c04fc170087119727119074e72445f2bb192b -.\" Author: Keller, Jacob E <jacob.e.keller@xxxxxxxxx> -.\" .\" SO_MAX_PACING_RATE (3.13) .\" commit 62748f32d501f5d3712a7c372bbb92abc7c62bc7 .\" Author: Eric Dumazet <edumazet@xxxxxxxxxx> @@ -869,6 +865,29 @@ Indicates that an unsigned 32-bit value ancillary message (cmsg) should be attached to received skbs indicating the number of packets dropped by the socket since its creation. .TP +.BR SO_SELECT_ERR_QUEUE " (since Linux 3.10)" +.\" commit 7d4c04fc170087119727119074e72445f2bb192b +.\" Author: Keller, Jacob E <jacob.e.keller@xxxxxxxxx> +Masks +.B POLLPRI +with +.B POLLERR +event return. This was created when +.B POLLERR +wake up depended on requesting +.B POLLIN +or +.BR POLLPRI . +After the commit 6e5d58fdc9bedd0255a8 ("skbuff: Fix not +waking applications when errors are enqueued"), introduced +in Linux 4.16, +.B POLLERR +wakes up normally. +The option has no effect in wake up behavior, all it does +is masking +.B POLLPRI +before poll returns. +.TP .B SO_SNDBUF Sets or gets the maximum socket send buffer in bytes. The kernel doubles this value (to allow space for bookkeeping overhead) -- 2.17.1
